Redbull Air Race Competition Entry

The Retro Racer.

By

Mike Robey

Introduction

I drew inspiration for this model from two sources. Firstly Hobbyking came out with their Durafly Slow Poke.

I then found out that the Slow Poke represented a whole genre of model aircraft much like the Ugly Stik.

I didn't want to design a Slow Poke but it reminded me of that 1930's low wing sports plane look and that I did want to emulate.  The second source of inspiration was a full sized race plane from the 1930's called the Comper Swift.

I liked the radial engine and the shape of the wings, tailplane and vertical stabilizer.

So I decided to mash the two ideas together to form a low wing 1930's racing plane with flying surfaces based on the Comper Swift and a cowled in radial engine at the front.  A short while into the design I watched the Flitetest video describing the Redbull Air Race Competition. Perfect timing! The result turned out pretty good if I don't say so myself.  There were some stumbling blocks along the way. For example a loose motor shaft which had to be removed so I could file a flat onto it after the plane was finished. However the most dramatic was my hand tremble becoming so severe that I thought my flying days were over. Fortunately that didn't turn out to be the case.

Design Goals

I planned to go nuts with 3D parts for this one to get it looking as cool as I could manage with a foam board airframe.  When I built and flew my first Bloody Baron, the wide speed range it was capable of impressed me as did the benign stall characteristics. Dan Sponholz had designed a brilliant little model with a fantastic aerofoil section. I have used it in many of my past designs and saw no reason not to employ it in this one.  Its only drawback is the CG needs to be quite forward so one needs to keep that in  mind when designing the fuselage.  I have noticed that wings with a wider chord tend to be more stable and generally better behaved than narrow chord wings so I did widen the chord a bit.  I assume that is why the Slow Pokes tend to have such a really wide chord. I wasn't going to go that wide but I did want to widen a tad.  I thought with the nose and 3D printed engine cylinders that getting the CG forward enough would be less of an issue because I figured all that 3D printed plastic in the nose was going to weigh quite a bit. Turns out I was wrong on that score which made things difficult when trying to place the battery to get the correct CG.  I started by designing the entire airframe in Sketchup as a 3D model.  I find this an advantage because I can get a good idea of what the finished aircraft will look like. Also I have a head start on all the 3D parts. After some Sketchup bashing I came up with the model shown in the image below.

After that I was off and running. I transfered the measurements manually from the 3D model in Sketchup to the 2D plans I was drawing in QCad.  Before I knew it I had a set of planes for the foam board parts.  Finishing the engine cylinders and all the other 3D printable models took quite a bit longer.  Overall the plane came together as expected. There were a couple of wrinkles which are explained in the build section but eventually my prototype Retro Racer was sitting on my build table ready for its maiden flight.

The Maiden

The weather forecast for maiden day had held at fine with light winds all week and then, on the day before maiden day, the forecast changed to rain and high winds. I went to bed the night before thinking the maiden would not happen as scheduled. I awoke on maiden day to an overcast sky with not a breath of wind. The maiden was on!  I was quite nervous about it. If you have read my previous article then you know I have a hand tremble which caused me to use extended control sticks on my transmitter. I had successfully completed a number of successful quadcopter flights but I only had two model plane flights under my belt at this point and I was still getting used to flying a model plane in this manner.  Add to that I had the CG close to where I wanted it but it was still a little further back (about 3mm) than I liked. Finally, if I destroyed my prototype on its maiden there would not be enough time to build a second one before the Redbull competition closed. All of that added up to a much higher than normal maiden anxiety level.

I arrived at the field and we got ready. I had several friends helping me. Carl Bradshaw who would chase me with his quad getting me some air to air video and Dan Weaver and Wayne Douglas taking turns to video the plane from the ground. The above image (image courtesy of Dave London) shows my partners in crime.  Carl seated on the left ready to chase me with his quad. Wayne also on the left (just to the right of Carl) and Dan is on the right getting ready to video using my camera.  You can also see me sitting ready with my modified transmitter sticks.  I positioned the plane ready, Carl's quad was fired up and ready with its camera recording, Dan had my camera recording. It was go time!. I slowly advanced the throttle from idle to full power and the plane just sat there not moving an inch.  Subsequent investigation revealed I had the motor running backwards.  Rookie mistake and one heck of an anti-climax! I would have sworn in a court of law that I had checked that during the build but clearly I had forgotten.  I had not brought any tools but with the help of people who did and Wayne, who's fingers were clearly much stronger and agile than  mine, we managed to get two motor wires swapped and were ready for another attempt.

On the second attempt things went well. There was a slight swing to the left easily countered with some right rudder and the Retro Racer lurched off the ground clawing its way skyward with ease. There was clearly plenty of power. I had forgotten to ease off the up elevator once the plane got moving which caused the dramatic lift off instead of a more graceful rising. I also forgot to ease off the right rudder I had applied on takeoff for a few laps of the field which made left turns a tad interesting. I was also over controlling in pitch. Mostly due to still getting used to my new stick arrangement and being really, really nervous. I realized I still had right rudder on and eased that off and got a handle on the over controlling in pitch and then things went pretty good. Carl was having trouble keeping up with the plane so I eased back to half throttle. For a draggy, open cockpit low winger, the Retro Racer was really moving along. Even at half throttle Carl was having trouble keeping pace. He's a clever fellow though and he started cutting corners turning before I turned so he would catchup as I zoomed past.  By the time it came for landing I was mentally exhausted. The approach was way too fast and I began the flare before the plane had slowed down sufficiently. I compensated up to a point but it still ended up with a nose over.  I can safely say the undercarriage got a good test.  The second flight was much better. My nerves had eased considerably now that the much feared unknown was revealed as being nothing to worry about. I also had video in the bank. Left turns go much better without right rudder so remembering to centre the rudder when required turns out to be a good thing. I was still a bit surprised by the speed so Carl was still struggling to catch up and my landing approach was still too fast but my flare was much better so the landing ended with the plane the right way up.  For those who are wondering Carl has much faster quads but the quad he was using was the one he had flown the longest and the only one he had used for chasing planes so we figured that was the wisest choice. Also he had never had a problem chasing my other planes and none of us were expecting the Retro Racer to be quite so fast.  If I had been thinking I should have tried a few stalls and got a feel for just how much I could safely slow the plane down. Unfortunately I only had that thought on the drive home. The link below is to the video I created from maiden day.

I went for a 1930's theme with the music so, if it annoys you then just mute the sound. I hope you enjoy it and huge thanks to Carl, Dan and Wayne for all their help.

Specifications

• Wingspan: 1096mm 

• Length: 900mm (including spinner)

• Flying Weight: 1190gms 

• Empty Weight: 980gms 

• Motor: Turnigy Air 3730-1000kv

• ESC: Turnigy Plush 40 amp 

• Prop: 10 x 5 APC

• Battery: 2200 mAh, 3S Turnigy Multistar

• Max Watts: 245 watts 

• Max Amps: 24 amps 

• Max Thrust: 1150 grams 

• Max C: 11

• Flight Time at Full Throttle: 5 minutes

3D Printed Parts

There are a fair number of 3D printed parts.  If you want your own Retro Racer and do not have a 3D printer then maybe there is a Flitetest member near you who has a 3D printer and doesn't mind printing them off for you. In fact maybe there should be a thread on the Flitetest forum for just that purpose (there may already be one I haven't checked).

Engine Cylinders

Print these with a raft. In spite of the pipes running from top to bottom they do not need any support. I glued my 3D bits together with CA glue.

Nose

I broke the nose into two parts so they could be printed without any support. The exit hole for the prop shaft is sized for a 40mm spinner with two degrees of down thrust and two degrees of right thrust.  The holes for the engine cylinders are intentionally sized slightly small. This is so each engine cylinder can be fitted via some gentle filing of the opening to ensure a tight fit.

Spinner

Pintokitkat's spinner is brilliant. He designed it to be 50mm in diameter. I have resized it to 40mm and then down to 30mm (for my Carbon Cub project). I did have to modify the top part to allow for a larger prop blade on this one. When assembled, two small servo screws hold the spinner top onto the base. I printed mine solid (i.e 100% infill).

Motor Mount

The motor mount has two degrees of side and down thrust built in and is aligned so the motor shaft will exit the nose piece dead centre.  The two T pins are slotted the the holes in the side of the motor mount to provide contact area with the fuselage side when gluing the motor mount onto the fuselage. The top piece fills the gap between the nose and the front deck. It should be part of the motor mount but I forgot and didn't realize until the motor mount was already glued in place. These should be printed solid for strength (i.e. 100% infill).

Control Horns

I usually print these solid as well. Obviously any suitable control horn could be used.

Battery Tray

To get the CG as far forward as I could the battery tray butts up against the motor mount and the battery goes in at the sharpest angle possible.  The half round bit is to allow  space between the tray and the rear of the spinning motor shaft.

Cockpit Coaming

I print this in TPU and glued it to the cockpit opening in the front deck using UHU Por glue.  

Front Deck

This is printed in 3 parts and then glued together. The fuselage plans show a foam board piece which is glued underneath to form a bottom.

Landing Gear Fuselage Mount

The piano wire landing gear legs fit into the slots on the inside faces of each part and then the two parts are bolted together.  The horizontal plates are to spread the shock loads of hard landings to avoid wear and tear on the foam board fuselage bottom. I used 3mm piano wire which has proved to be strong enough.

Battery Hatch Holder

This slots into the foam board and glued to the bottom of the front deck so that it pokes out the bottom. When the front deck is fitted to the fuselage this piece slots through a hole in the fuselage top and is held in place with a barbecue skewer cut to size which slides through one side of the fuselage, into the hole in the bottom of the hatch holder and out the other side of the fuselage. The holes in the fuselage sides are not shown on the plans because this is best done in situ, after fuselage completion, to get a good tight fit.

Headrest

This part is glued on top of the poster board rear deck.

Rigging Wire Attach Points

These poke through the outboard section of each wing half and are glued in place. The rigging wires are then attached via the holes. As the rigging wires are purely cosmetic so I simply used elastic cord for the flying wires.  I originally intended to route the rigging wires through a fifth point attached to the top of the front deck before my brain caught up and I realized the front deck is the battery hatch and so has to be removable.  You can see how I got around that issue in the build section below.

Tail Skid

I have used this tail skid on most of my models. Does the job well.

Windscreen

Technically its actually a windscreen frame. I suppose you could glue some transparent plastic on the inside but I prefer to imagine  my pilot keeps his windscreen really, really clean. Its so clean you can't even see the glass.

Wheels

For all my wheels, the hubs are printed in ABS or PLA with 10% infill and the tyres are printed in TPU with 5% infill.  That amount of infill with the TPU gives just the right amount of flex in the tyres.  These were actually designed for my Carbon Cub project but I liked the forgiveness these tyres were going to provide on my first few landings with the Retro racer so I decided to go with these.

I did however design some more wheels more authentic to the 1930's. The Sketchup model and the actual wheels are shown in the above images.

The Build

The following build guide assumes a basic familiarity with FT style builds. If you are new to FT building the watch one of Flitetest's excellent build videos before proceeding. The build guide for the FT-Spitfire is probably the closest to this one.  Here is a video overview of the design and build process for the Retro Racer:

Normally I start with the foam board fuselage assembly but with this one I started with assembling the 3D printed nose section.

The image above shows the cylinder and nose parts fresh off the 3D printer. I did use pictures of radial engines when designing the cylinders but dumbed it down a lot to reduce the task of modeling them in Sketchup.  After all this is a foam board model of a non-existent aircraft so I didn't need to get carried away.

I did design the cylinders with a different front and back purely for aesthetics. I glued them together with CA glue which I find provides a really strong bond with 3D printed parts.

The hole in the nose for each cylinder is carefully filed until the cylinder friction fits into place. They will be glued in place after painting but having a tight fit makes it much easier for them to all stay in the right place while applying the glue.

The cylinders were spray painted silver and the nose was hand painted. I planned to use blue wing tape for most of the trim so the blue paint had to be as close as possible match to the blue of the wing tape.

Next the cylinders were fitted to the nose section and glued in place.  Keep in mind that in a nose over or a cartwheel (let's hope not but better safe than sorry) the cylinders are going to take a hit so don't spare the CA. I tacked them in place on the outside and then applied lots of glue on the inside of the nose.

The next step was to assemble the undercarriage. I designed a template for bending the piano wire (included in the plans).  I must be getting old and feeble because I really struggled to bend the 3mm piano wire whereas I used to be able to bend it no problem. Isn't getting old so much fun?  Your wire doesn't have to match the template perfectly (mine didn't). The main thing is it fits into the slots on the 3D printed undercarriage mounts and the wheel axles are level with each other.

Once the mounts are bolted together I was ready to commence fuselage construction. I suppose I could have epoxied the two mount parts together instead of bolting but I figure you can never have an undercarriage mount that is too strong.

The above image shows all the fuselage parts. The pilot is from Thingiverse:

https://www.thingiverse.com/thing:1750200

From memory I had to scale him up to get him big enough. I tried to import the STL file into Sketchup but there are too many data points. He has a mustache which I chose to ignore this time when painting him up.  Given my shaky hands I doubt I can paint another one so, if I ever need another Manfred, I will have to ask a friend to paint him for me. Its an excellent pilot figure though.

Now I was ready to start on the fuselage.  Just like all Flitetest fuselage's we start by gluing one side making sure its perpendicular to the fuselage bottom.

Then we glue the other fuselage side to the bottom. Again ensuring it is perpendicular to the fuselage bottom.

Next I glued the fuselage top in place. Once the hot glue was applied, I quickly turned the fuselage upside down and pressed it against the work table while gently sliding it back and forth while the hot glue cooled. This ensures a good, close fit. Similarly the rear fuselage bottom was then glued into place using the same process.

The rear fuselage coaming doesn't go all the way to the tail plane so there was a small piece to glue onto the fuselage top to fill the gap. Same process of applying the glue, flipping the fuselage over and pressing down on the workbench until the glue has cooled was used to ensure a good flat join.

The next step was to glue the two rear fuselage formers in place. Note the vertical slots.  A horizontal piece is glued into the slots to give the poster board a little more support.  Speaking of the horizontal piece...

The horizontal piece was glued in place next. Its easy to get confused about which way up this piece goes. Note the pencil arrow drawn on the front of the piece to ensure I got that correct.

I was a bit of a drongo (Aussie slang for idiot) and I forgot to photograph the assembly of the front deck.  The hatch fastener was glued in  place onto the foam board front deck bottom. The three printed pieces of the front deck were then placed onto the foam board bottom and tacked together with CA when positioned correctly. Then the base was removed and more CA applied to each join.  The cockpit coaming was then glued in place. I used UHU Por for that. I used some pegs to hold it in  place and left it over night to ensure a good strong bond. Finally the foam board base is glued to the top.

The next step was to bolt the motor to the motor mount. This is a really good motor which can extract pretty much as much power as you can get from a 3S, 2200 mAh battery and yet still have a reasonable flight time (see above specifications). Its drawback is you have to reverse the motor shaft which annoys the crap out of me. Very few people front mount their motors these days so I don't know why they don't have the shaft the right way around to start with. However, once the shaft is reversed, its a great little motor so I guess I shouldn't complain too much.

Next the side plates were CA glued to the motor mount. These are simply to provide more surface area when gluing the motor mount to the fuselage. Not sure if its actually needed as I have never had a mount that didn't employ side plates come loose but it makes me feel better.

The next step was to glue the motor mount to the fuselage. I used epoxy for this. Its always fun trying to figure out how to keep the model in place while the glue dries.  You cannot see it properly in the image above but the steel ruler is being weighted down with a cordless power drill. The G-clamps are ensuring the side plates are pushed solidly against the fuselage sides while the glue is drying.

Normally I paint all the 3D printed parts but it occurred to me that the front deck has no compound curves and the wing tape went on easily and looks much better than if I had painted it.  Also it was time to glue Manfred and the windscreen into place with UHU Por. The windscreen was hand painted.

The next step was to cover the poster board for the rear deck in wing tape and then glue it in place. After test fitting the rear deck was glued to the horizontal centre piece and then each side was glued down in the normal manner for Flitetest builds.  Its important to ensure that the amount of poster board overlapping the fuselage sides is the same on both sides so I trimmed one side a bit to balance that up. Finally the headrest was covered in wing tape and glued to the rear deck using UHU Por glue.

Next wing tape trim was added to the forward fuselage to match the wing tape on the rear fuselage.

As with all Flitetest designs the slots for the wing were not fully cut out when cutting out the fuselage. This was to ensure the fuselage sides didn't partially collapse when pushing down on them during fuselage construction. Hence the next step was to cut out the wing slots fully ready for the wing installation. Of course we need a wing to install so the next thing to do is...

Yep you guessed it. The wing is next. Note my original trim is pretty minimal. My daughter, who has a bit of a flare for these things, expressed her disgust at my minimal effort and encouraged me to do a lot better. I am very glad I took her advice.  As you can see from the image above the wing is stock standard Flitetest with a foam spar which slots into the wing bottom before folding and gluing.  There is one new step which is to glue the attachment points for the flying wires. They can just be seen sitting on the wing bottom.

The flying wire attachment points were then positioned in the wing and glued in place.  There was an unexpected consequence of this which you will see later.

The above image shows the flying wire attachment points poking out through the wing top after they were glued in place.

In typical Flitetest fashion the wing spar halves are folded and glued together. Its important to wipe any excess glue from the spar bottoms and where the two spar halves overlap each other.  This is to avoid a poor fit when gluing the spar halves to the wing bottom and when gluing the wing halves together.

The next step was to glue each spar half into its corresponding wing half. Its important to test fit first to ensure each spar half is going to easily pop into place when glue is applied.

Next hot glue was applied to the leading edge and spar top and the wing tops folded over and held in place for a good five minutes to ensure the hot glue has fully set. Then hot glue was applied to the trailing edge of the wing bottom and held down again for another five minutes. Five minutes is overkill but there is nothing worse than picking up a wing and realizing that the wing top has started to part from the wing bottom because they were not held in place long enough (ask me how I know!).

The wing halves were then glued together and...

The dihedral jig was used to ensure 2.5 degrees of dihedral for each wing half.  Note I made a bottom for the dihedral jig to ensure it doesn't fall over.

I like to leave the wing to dry for another good five minutes to ensure everything is rock solid before proceeding to the next step.

Fibreglass re-enforced tape was then applied to the wing joint making sure to cut the tape away from the servo lead hole. I have forgotten this step once and the pain it caused means I will never forget again. Its so much fun to cut the tape away after the wing is installed into the fuselage. Its doable but its not fun.

The next step was to slide the wing into the fuselage wing slot. It was at this point an unexpected problem arose.  I could not slide the wing through the wing slot because the flying wire attachment points wouldn't fit. After a bit of grumpy mumbling to myself( I swear I just said golly gosh. No really!).  I realized the only way forward was to cut little slots into the wing slots on each side of the fuselage to allow the flying wire attachment points on the wing to slide through.

I kept the foam bits that I cut out and glued them back once the wing was in place. This, other than trying to get the CG in the right place, was the only unforeseen issue (Okay I suppose I also had the motor running backwards so maybe not the only unforeseen thing).  The wing was centred by measuring along the leading and trailing edges on each side and gently nudging the wing until it was centred nicely.  Finally hot glue was applied to each wing/fuselage join top and bottom making sure to wipe away the excess glue with a scrap piece of foam board.

The next thing to do was glue the vertical stabilizer to the tailplane making sure they were perpendicular to each other. Also I checked the elevator and rudder had full movement and found the rudder was blocked a bit so I widened the hole in the rudder where the elevator goes through to ensure full and free movement of both control surfaces.

There wasn't enough clearance between the fuselage bottom and the wing for the undercarriage mount to fit. I did foresee this but I decided it would be better to cut away the required area on the wing at this point rather than trying to allow for this on the plans.  Also I had to cut away the fuselage bottom a bit to allow for the bolts holding the two halves of the undercarriage mount together.  This was easily achieved.  From memory I think only the wing bottom needed cutting but not the wing top.  Use a sharp blade as you are cutting through the hot glue that was applied to the wing leading edge.

The undercarriage mount was then glued into place. I used epoxy for this purpose. Note the hole in the fuselage bottom behind the motor mount. I left this open in case I needed to access the motor wires in the final stages of construction. I closed it up with a piece of foam board when I was sure I was done. The irony is that I forgot to check the motor direction when I plugged the ESC into the motor. You know what they say, "The best laid plans...".

The tail feathers were then  positioned and adjusted until they lined up perfectly with the wing and then hot glued in place. The tail skid was then glued on using epoxy.

The nose was then  glued onto the motor mount. I used UHU Por so that it can be easily cut and removed in the future if needed. Also note the creative juggling of tool boxes and drills to hold the model vertical while the glue was drying.

I used string with a lead sinker to get the servo wires in place. The elevator and rudder servo were easy. The aileron servos were a bit more fiddly but easier than expected. The next step was to bind and hook up the receiver and then I was in the home stretch. Note also the flying wires were simply elastic cord tied to the corresponding flying wire attachment points and sticky taped in place in the correct position in the fuselage. Simple and neat.

The final step was to figure out how to get the battery forward enough to get the required CG. It was a definite challenge. The CG should be dead centre on top of the wing spar. Mine is 3 or 4mm behind that but, from the experience gained from the maiden flights, its forward enough. I suppose I could have used lead to get the CG exactly where I wanted it but that seemed like blasphemy so I decided to settle for this. I also had to trim the bottom of the forward to deck to allow the battery to poke up into the cavity under the front deck.

The above image shows how the Retro Racer looks with the 1930's period wheels instead of the balloon tyres. Definitely more authentic looking but I think I am sticking to the balloon wheels.

Data Files

The pdf files for the plans can be found here: pdfFiles.zip

The dxf files for the plans can be found here: dxfFiles.zip

The sketchup files for the 3D printed parts can be found here: SketchupFiles.zip

The STL files for the 3D printed parts can be found here: STLFiles.zip

Conclusion

The Retro Racer was a challenging from both the overall design and especially from the 3D modeling for the 3D printed parts. It also had its fair share of unexpected issues (e.g. the motor shaft) and I, personally had issues (i.e. trembling hands temporarily stopping me flying). The maiden was much more stressful than normal but I am very happy with the result. The plane exhibits good flying characteristics and looks great on the ground and in the air.  The maiden, though nerve wracking, was also very special to me because it involved making memorable moments with like  minded friends.  I am grateful for their help and for every second of the design, build and maiden flight that I experienced. On top of that I have a cool looking  model that came from my own mind and hands (shaky though they are). Now all this is done and dusted its time to get back to the Carbon Cub. If all goes well, more on that later.